Knowledge-Based System for Conceptual Schema Conversion on a Multi-Model Database Machine

Abstract

The design of a “good” conceptual schema for a multi-model database machine or application is often the object of an iterative process that is performed usually manually and in an ad-hoc manner. With the advent of knowledge-based systems and rule-based programming, it has become feasible to encapsulate a substantial amount of the database administrator’s knowledge into a system capable of performing normalization operations and making database design decisions. The interactive knowledge-based system described in this paper attempts to automatize the normalization process of a E/R conceptual schema and to produce an equivalent “better” normalized schema. The ultimate goal is to produce a “proper” conceptual schema i.e. a conceptual schema that yields lossless, dependency-preserving and normalized external schemas.